Glaucoma is one of the leading causes of treatable blindness in the world. However, the manner and rate of disease progression varies markedly between patients, and so predicting future progression and hence the prognosis for an individual is challenging. Much of the previous work in this field has concentrated on determining whether a patient with certain risk factors will develop glaucoma. There has been less attention paid to the likelihood that a patient will suffer significant loss of vision as a result of their glaucoma, or how soon vision loss may manifest. The overall goal of this project is to improve the ability to predict the future course of an individual's disease, and so enable the clinician to adapt their management strategy accordingly. Better knowledge of glaucomatous pathophysiology will also be gained, with meaningful implications for developing new methods of diagnosis and treatment strategies. The primary tool used in this project will be longitudinal data collected biannually from over 250 subjects with early or suspected glaucoma, and 50 normal subjects. For the majority of the glaucoma subjects, up to ten years of prior information is available and will also be used. The first part of the project aims to improve understanding of the structure-function relationship in glaucoma. Three non-competing hypotheses will be tested that seek to explain the weak observed correlation between functional change (to the patient's visual field) and structural change (to the optic nerve head and/or retinal nerve fiber layer). The first experiment will determine the true underlying correlation after taking into account the variability inherent in the testing procedures. The second experiment will compare changes to the optic nerve head surface topography against subsurface changes and changes to the retinal nerve fiber layer, to determine whether there is temporal dissociation caused by surface change not driven by a loss of neural tissue. The third experiment will use a sampling-limited test of retinal ganglion cell density, to determine whether there is temporal dissociation caused by dysfunctional retinal ganglion cells that are still present in the retina but have altered response characteristics. Results from these three experiments will provide new insights into the disease process, with possible implications for future development of testing and treatment strategies. The second part of the project aims to use this new information and advanced analysis techniques (including regression trees) to better predict the future rate of functional change on a per-eye basis. First, subsequent change will be predicted based on test results at one time point. Second, improved predictions will be made when a series of test results are available. The fundamental motivation for this project is to enable a clinician to prevent severe visual disability or blindness in an ocular hypertensive or glaucoma patient, by identifying a rapid progression rate or a high likelihood for rapid progression at the earliest stages of the disease.